Modelado geométrico personalizado de la córnea humana y su aplicación a la detección de ectasias corneales

Abstract

[SPA] La córnea es una estructura biológica viva cuya arquitectura presenta una morfología singular, ya sea en un estado natural o patológico. Esta singularidad ha sido caracterizada a lo largo de toda la historia en el campo de la oftalmología y la óptica a través de la generación de modelos genéricos o de modelos personalizados de la córnea humana. Hoy en día, el desarrollo de nuevas tecnologías permite caracterizar la morfología corneal a partir de los denominados equipos topográficos; estos equipos aportan una caracterización personalizada de índole cualitativa y cuantitativa al médico oftalmólogo. Sin embargo, los sistemas de diagnóstico de las patologías corneales están basados en unos índices de valoración de la irregularidad de las superficies corneales que son calculados a partir de algoritmos específicos internos para cada topógrafo corneal y de los cuales se desconoce su programación. Por este motivo en esta tesis doctoral se establece un nuevo procedimiento fundamentado en la geometría computacional para obtener un modelo sólido 3D personalizado in vivo de la córnea humana utilizando herramientas de Diseño Geométrico Asistido por Ordenador. Este modelo virtual reconstruye fidedignamente las superficies de la cara anterior y posterior de la córnea, a partir de unos datos aportados por los topógrafos corneales denominados datos en bruto (sin ningún trato mediante algoritmo) tanto para los ojos de pacientes sanos como para los ojos de pacientes diagnosticados con la patología ectásica más común, el queratocono. A partir del nuevo modelo sólido obtenido, se definen unos índices de caracterización de la morfología corneal basados en variables geométricas, los cuales pueden ser utilizados como unos nuevos índices de diagnóstico de la patología ectásica objeto de estudio debido a que presentan una elevada sensibilidad y especificidad para su diagnóstico. [ENG] The cornea is a living biological structure whose architecture has a unique morphology, either in a natural or diseased condition. This uniqueness has been characterized throughout all history in the field of ophthalmology and optics through the generation of generic or customized models of human cornea. Today, the development of new technologies leads to characterize the corneal morphology from the so‐called topographic devices; these devices provide a personalized qualitative and quantitative characterization of its nature for the ophthalmologist. However corneal pathological diagnosis systems are based on indicators of the irregularity of the corneal surfaces, which are calculated from specific internal algorithms for each corneal topographer and whose programming is unknown. For that reason, this doctoral thesis establishes a new procedure based on computational geometry to obtain a 3D solid model, personalized and in vivo of the human cornea by using Computer Aided Geometrical Design tools. This virtual model represents accurately both the anterior and posterior corneal surfaces from a set of raw data (without any algorithm treatment) provided by the corneal topographers for both healthy corneas and corneas with the most common ectasic disease, the keratoconus. The new solid model obtained is later analyzed to define a set of indices that enable the characterization of the corneal morphology and that are based on geometric variables. These indices can be used as new indicators for the diagnosis of the keratoconus disease due to their high sensibility and specificity.

[ENG] The cornea is a living biological structure whose architecture has a unique morphology, either in a natural or diseased condition. This uniqueness has been characterized throughout all history in the field of ophthalmology and optics through the generation of generic or customized models of human cornea. Today, the development of new technologies leads to characterize the corneal morphology from the so‐called topographic devices; these devices provide a personalized qualitative and quantitative characterization of its nature for the ophthalmologist. However corneal pathological diagnosis systems are based on indicators of the irregularity of the corneal surfaces, which are calculated from specific internal algorithms for each corneal topographer and whose programming is unknown. For that reason, this doctoral thesis establishes a new procedure based on computational geometry to obtain a 3D solid model, personalized and in vivo of the human cornea by using Computer Aided Geometrical Design tools. This virtual model represents accurately both the anterior and posterior corneal surfaces from a set of raw data (without any algorithm treatment) provided by the corneal topographers for both healthy corneas and corneas with the most common ectasic disease, the keratoconus. The new solid model obtained is later analyzed to define a set of indices that enable the characterization of the corneal morphology and that are based on geometric variables. These indices can be used as new indicators for the diagnosis of the keratoconus disease due to their high sensibility and specificity.